guizang-ppt

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill operates as a local template generator, producing static HTML files without requiring server-side execution or sensitive data access. The provided workflow and layouts are consistent with legitimate creative use cases.\n- [EXTERNAL_DOWNLOADS]: The generated PPT template references styling assets from trusted, well-known providers. Specifically, it loads font families via Google Fonts (fonts.googleapis.com) and the Lucide icon library via UNPKG (unpkg.com). These are standard industry practices for web-based presentations.\n- [COMMAND_EXECUTION]: To initialize projects, the skill instructs the agent to perform basic filesystem operations such as creating image directories (mkdir) and copying the core template file (cp). These actions are scoped to the project directory and present no privilege escalation risk.\n- [INDIRECT_PROMPT_INJECTION]: The skill is designed to transform user-provided outlines and content into a presentation.\n
  • Ingestion points: User-supplied text, structure, and image paths ingested during the PPT generation process.\n
  • Boundary markers: Not present, as the content is directly interpolated into HTML templates.\n
  • Capability inventory: The agent can write to the local filesystem (index.html) and suggest local browser viewing (open).\n
  • Sanitization: Content is not actively sanitized, but the risk is mitigated by the fact that the output is a static local HTML file meant for presentation preview, rather than for execution in a privileged or network-exposed environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 09:56 PM
Security Audit — agent-trust-hub — guizang-ppt